Ignore the output from the several calls gc() used to maximize available memory!

##             used   (Mb) gc trigger   (Mb)   max used   (Mb)
## Ncells   2300078  122.9    4203792  224.6    3156587  168.6
## Vcells 402947783 3074.3 1137859298 8681.2 1261594852 9625.3
##             used   (Mb) gc trigger   (Mb)   max used   (Mb)
## Ncells   1778745   95.0    4203792  224.6    3156587  168.6
## Vcells 187129650 1427.7  910287439 6945.0 1261594852 9625.3
##             used   (Mb) gc trigger   (Mb)   max used   (Mb)
## Ncells   1527007   81.6    4207162  224.7    3157514  168.7
## Vcells 271386383 2070.6  937809599 7155.0 1261781468 9626.7
##             used   (Mb) gc trigger   (Mb)   max used   (Mb)
## Ncells   1529618   81.7    4207162  224.7    3157514  168.7
## Vcells 271650518 2072.6  750247680 5724.0 1261781468 9626.7
##            used (Mb) gc trigger   (Mb)   max used    (Mb)
## Ncells  1540266 82.3  106214847 5672.5  259313588 13848.9
## Vcells 10282611 78.5 1310018111 9994.7 1609985633 12283.3

Maps of relative county excess mortality as a % (excess / expected) by COVID wave and Metro status

  • Large metro = NCHS categories 1 + 2

  • Small/Medium Metro = NCHS category 3

  • All Metro = NCH categories 1, 2, + 3

  • Non-metro = 4

Heatmaps of relative county-month excess mortality as a % (excess / expected) by census region and Metro status

  • Large metro = NCHS categories 1 + 2

  • Small/Medium Metro = NCHS category 3

  • All Metro = NCH categories 1, 2, + 3

  • Non-metro = 4

Rows (counties) ordered by peak month + second-highest month

Rows (counties) ordered by unsupervised clustering

NB: Haven’t been able to generate a visually patterned order this way, so not outputting it.